Translating nature into medicine at Enveda Biosciences with machine learning and Bruker’s trapped ion mobility tandem mass spectrometry (TIMS MS/MS)

Introduction

Enveda Biosciences is a biotechnology company using a combination of breakthroughs in metabolomics, machine learning, automation and biological assays to revolutionize natural product drug discovery. By harnessing the world’s largest and most diverse biologically annotated dataset of plant chemistry, Enveda is looking to discover the next generation of impactful small molecule therapeutics.

Enveda Biosciences was founded in 2019 and has relied on the Bruker timsTOF technology since 2021 to investigate the chemical composition of complex samples collected from natural sources. The company’s vision is to lead a shift from screening to informed database searching as the first step in drug discovery. The timsTOF Pro has become a core element of the technology suite to identify, annotate and utilize the structure and function of life’s chemistry.

Natural products – a novel method of drug discovery

Historically, phytochemicals (molecules derived from plants) have played a pivotal role in the field of drug discovery, owing to their vast array of chemical attributes and biological functions. However, natural products only represent a small percentage of new small molecule drugs brought to market.

For Enveda, molecules from plants represent an enormous untapped library of drug-like chemicals that can be developed into novel small molecule drugs, though current methods to identify these novel molecules from natural sources are challenging and labor intensive. Enveda’s approach is to build and utilize a unique database of natural compounds, starting with plants, leading to rapid discovery of previously unidentified natural products to generate new medicines for patients suffering from a range of diseases.

An industrial-scale platform

Having to isolate and study potentially therapeutic molecules one at a time has made the conventional approach to natural product drug discovery time consuming, expensive and prone to failure. Most natural products are unknown to science, meaning that their identity, particularly their chemical structure, is a mystery.

To overcome these challenges, Enveda has revolutionized the process through innovative, state-of-the-art technologies and techniques, demonstrating that drug discovery can be initiated by an informed database search – querying chemical structure, biological activity, and organ distribution ­– to rapidly search for new leads.

First using liquid chromatography (LC) to separate complex mixtures into fractions, Enveda sends a portion of each fraction though the timsTOF Pro 2 for structure prediction, while the other portion of the fraction is tested in dozens of different biological assays to determine activity. Statistical models are used to deconvolute the data and the molecules with the right combination of beneficial activity, low toxicity, and amenable structures are then prioritized for isolation.

Bioactive components of natural products tend to be present in low abundance, making it difficult to isolate large enough quantities to perform structure analysis using nuclear magnetic resonance (NMR). To overcome this challenge, Enveda has pioneered new ways of analyzing the data produced by high resolution MS using machine learning. The key to identifying natural product structures is the combination of clean MS/MS fragmentation patterns and ion mobility data generated from Bruker’s timsTOF Pro 2 mass spectrometers.

“The challenge in the industry is no longer the acquisition of MS/MS spectra itself but the rate of acquisition and the quality of the spectra. This is where the timsTOF platform has set itself apart from its competitors – it performs well with both and provides a single-point encoding of a molecule’s shape in the form of its collisional cross section (CCS), which can be utilized to link molecules between acquisitions even when separations are varied.” Pelle Simpson, Senior Scientist, Enveda Biosciences."

“We are incredibly proud of our work, but this is just the beginning. We will continue building the largest metabolomics dataset, purpose-built for machine learning, and applying active learning strategies to help us identify and characterize the mass spectra whose identity is most likely to improve our drug development programs.” August Allen, Chief Technology Officer, Enveda Biosciences."